易于使用的计算机

• 终极目标

  • make the system easy to use

• 计算机系统需要具备哪些性质才是易于使用的

  • 同一台机器上能够同时运行多个程序
  • 对于编程者来说，不用花太多的信息在内存上
  • 程序运行的效率尽可能高
  • 关机之后信息能保存
  • ······

• 操作系统
  OS的存在主要是为了使计算机系统易于使用。为了达到这个目标，OS需要完成以下三方面的工作

  • 虚拟化（virtualization）
  • 并行（concurrency）
  • 持久化（persistence）

虚拟化

虚拟化可以分成两部分

• CPU的虚拟化
• 内存的虚拟化

虚拟化CPU

目标

让用户觉得自己的单核CPU电脑能够同时运行多个程序

Mechanism

• 两个问题
  
  • 当前哪一个进程获得CPU（scheduling policy CPU调度）
  • 进程之间怎么切换

进程切换

• Direct Execution Protocol
  
  • 
  • problems:
    
    • 不能能保证program一定不执行对系统有害的操作
    • program可以一直不把系统的控制权还给OS

• Limited Direct Execution Protocol

  • 
  • problems
    
    • 虽然能保证program不执行对系统有害的操作，但是无法保证program一定会把系统的控制权还给OS

• Limited Direct Execution Protocol (Timer Interrupt)

  • Limited Direct Execution Protocol (Timer Interrupt)
  • 
  • 完美解决上面遇到的两个问题

进程调度

在进程切换的时候，OS应该选择哪一个进程占用CPU才是合理且高效的呢？

Policies

• Assumptions

  • Each job runs for the same amount of time.
  • All jobs arrive at the same time.
  • Once started, each job runs to completion.
  • All jobs only use the CPU (i.e., they perform no I/O)
  • The run-time of each job is known.

• Targets：

  • 
  • 

• Scheduling policies

First In, First Out (FIFO)

• 如果假设1不成立

Shortest Job First (SJF)

• 在假设1不成立的时候依然能够很好的工作
• 如果假设2不成立

Shortest Time-to-Completion First (STCF)

• 在假设1，2，3不成立的时候依然能够很好的工作
• 但是对于现代的分时操作系统来说，除了turnaround之外，响应时间或许是一个更加重要的指标

Round Robin（RR）

• 在假设1，2，3均不成立的时候，对response time友好，但是对于turnaround不友好

• 如果假设4不成立

  • 当当前任务执行I/O时，将CPU让给其他进程

多级反馈队列

多级反馈队列是一种实际运用相对较多的调度算法，在Windows NT内核中都有用到，我们单独拿出来讲

• The Multi-Level Feedback Queue(MLFQ)

  • Assumptions

    • 假设1，2，3，4，5均不成立

  • Targets

    • optimize turnaround time
    • minimize response time

  • Basic Rules

    • Rule 1: If Priority(A) > Priority(B), A runs (B doesn’t).
    • Rule 2: If Priority(A) = Priority(B), A & B run in RR.

  • Attempt #1: How To Change Priority

    • Rule 3: When a job enters the system, it is placed at the highest priority (the topmost queue)
    • Rule 4a: If a job uses up an entire time slice while running, its priority is reduced (i.e., it moves down one queue)

    • Rule 4b: If a job gives up the CPU before the time slice is up, it stays at the same priority level.

    • Problems:

      • starvation: if there are “too many” interactive jobs in the system, they will combine to consume all CPU time, and thus long-running jobs will never receive any CPU time
      • game the scheduler: before the time slice is over, issue an I/O operation (to some file you don’t care about) and thus relinquish the CPU
      • change its behavior: a program may translate to a phase of interactivity and never be treated as a interactive job

  • Attempt #2: The Priority Boost

    • Rule 5: After some time period S, move all the jobs in the system to the topmost queue.

    • Problems:

      • game the scheduler: before the time slice is over, issue an I/O operation (to some file you don’t care about) and thus relinquish the CPU

  • Attempt #3: Better Accounting

    • Rule 4: Once a job uses up its time allotment at a given level (regardless of how many times it has given up the CPU), its priority is reduced (i.e., it moves down one queue)

  • final solution

    • Rule 1: If Priority(A) > Priority(B), A runs (B doesn’t).
    • Rule 2: If Priority(A) = Priority(B), A & B run in RR.
    • Rule 3: When a job enters the system, it is placed at the highest priority (the topmost queue).
    • Rule 4: Once a job uses up its time allotment at a given level (regardless of how many times it has given up the CPU), its priority is reduced (i.e., it moves down one queue).
    • Rule 5: After some time period S, move all the jobs in the system to the topmost queue.

虚拟化内存

• 为什么我们需要虚拟化内存?
  

  Mostly ease of use: the OS will give each program the view that it has a large contiguous address space to put its code and data into; thus, as a programmer, you never have to worry about things like “where should I store this variable?” because the virtual address space of the program is large and has lots of room for that sort of thing. Life, for a programmer, becomes much more tricky if you have to worry about fitting all of your code data into a small, crowded memory—《Operating System: There easy pieces》

目标

• transparency
• efficiency
• protection

Mechanism

• Address Translation（）
  
  • Dynamic (Hardware-based) Relocation
    
    • Dynamic Relocation 基于硬件，在程序运行时发生，程序在中断之后，可以把虚拟地址映射到另外一个与之前不同的物理地址空间，并且提供进程之间的保护
  • static(software-based) relocation
    
    • static relocation 基于软件，在程序运行之前完成，不支持在运行时虚拟地址的重新映射，不提供进程之间的保护

• 

• 

Policies

• Naive solution

  • base and bound registers

  • physical address = virtual address + base

  • problems:

    • there is a big chunk of “free” space right in the middle, between the stack and the heap.
    • 

• Segmentation

  • instead of having just one base and bounds pair in our MMU, why not have a base and bounds pair per-logical segment of the address space
  • 

  • address translation procedure

    • 
    • 

  • problems:

    • external fragmentation
    • The only real solution is to avoid the problem altogether, by never allocating memory in variable-sized chunks

• paging
  为了解决分段带来的外部fragmentation问题，分页技术把内存分成大小相同的页

  • 
  • The simplest form is called a linear page table, which is just an array. The OS indexes the array by the virtual page number (VPN), and looks up the page-table entry (PTE) at that index in order to find the desired physical frame number (PFN)
  • address translation procedure
    
    • 
  • problems:
    
    • efficiency
      
      • time: 访问两次内存
      • space: 每个进程一个页表，占用太多的内存空间

• speed-up page table(TLB: translation-lookaside buffer)

  • A TLB is part of the chip’s memory-management unit (MMU), and is simply a hardware cache of popular virtual-to-physical address translations; thus, a better name would be an address-translation cache.

  • translation procedure with TLB

    • 

  • problems:

    • TLB Miss
      
      • TLB miss handler
        
        • depending on how a trap or exception was caused, the hardware must save a different PC when trapping into the OS, in order to resume properly when the time to do so arrives
      • avoid infinite chain of TLB misses
        
        • keep TLB miss handlers in physical memory
        • reserve some entries in the TLB for permanently-valid translations
  • Context Switches
    
    • address space identifier (ASID)
  • Replacement Policy
    
    • least-recently-used
    • swapping
  • how would you access cache?
    
    • virtually indexed cache or physically-indexed cache?

• save memory space

  • Bigger Pages
    
    • problems:
      
      • big pages lead to waste within each page(internal fragmentation)
  • Hybrid Approach: Paging and Segments
    
    • use the base not to point to the segment itself but rather to hold the physical address of the page table of that segment
    • unallocated pages between the stack and the heap no longer take up space in a page table
    • translation procedure
      
      • 
    • problems:
      
      • have a large but sparsely-used heap, for example, we can still end up with a lot of page table waste
      • page tables now can be of arbitrary size
  • Multi-level Page Tables
    
    • page table of page table
      
      • 
      • The page directory thus either can be used to tell you where a page of the page table is, or that the entire page of the page table contains no valid pages
    • translation procedure
      
      • 
      • 

• Inverted Page Tables

  • keep a single page table that has an entry for each physical page of the system. The entry tells us which process is using this page, and which virtual page of that process maps to this physical page.

• Swapping the Page Tables to Disk

  • place such page tables in kernel virtual memory, thereby allowing the system to swap some of these page tables to disk when memory pressure gets a little tight

• swapping

  • swapping技术的存在同样是服务于primal goal
  • 在PTE中有一个present bit字段，如果这个字段为0，表示当前页不在内存中，也就是说产生了一个page fault
  • 

  • 

  • Replacement algorithm

    • goal
      
      • minimize
    • The Optimal Replacement Policy
      
      • furthest in the future
      • 
    • A Simple Policy: FIFO
    • Another Simple Policy: Random
    • Using History: LRU

    • workloads affect choice of policy

      • 

      • 

    • implementing LRU

      • perfect LRU
        
        • as the number of pages in a system grows, scanning a huge array of times just to find the absolute least-recently used page is prohibitively expensive

      • approximating LRU（clock algorithm）

        When a replacement must occur, the OS checks if the currently-pointed to page P has a use bit of 1 or 0. If 1, this implies that page P was recently used and thus is not a good candidate for replacement. Thus, the use bit for P set to 0 (cleared), and the clock hand is incremented to the next page (P + 1). The algorithm continues until it finds a use bit that is set to 0, implying this page has not been recently used

        • 

并发

• 线程
  
  • a new abstraction for a single running process: that of a thread
  • a multi-threaded program has more than one point of execution (i.e., multiple PCs, each of which is being fetched and executed from)
  • context switch（TCBs）
  • 
• KEY CONCURRENCY TERMS
  
  • 

• goals

  • 引入线程的目的是希望我们的程序fast and correct
  • 怎么保证correct？
    
    • mutual exclusion
  • 怎么实现mutual exclusion？
    
    • 同步原语（synchronization primitives）

• 同步原语

  • 同步原语是什么？
    
    • 像锁、信号量、条件变量等都是同步原语
    • 
  • 锁、条件变量和信号量之间的区别和联系是怎样的？
    
    • 锁是用来实现互斥访问的基本同步原语
    • Semaphores and condition variables build on top of the mutual exclusion provide by locks and are used for providing synchronized access to shared resources. They can be used for similar purposes.
    • A condition variable is generally used to avoid busy waiting
    • 信号量能够同时完成锁和条件变量的功能
      
      • binary semaphore并且初始值是1的时候相当于锁
      • binary semaphore并且初始值是0的时候相当于条件变量
      • 虽然信号量能完成锁和条件变量能完成的事，但是二者之间的联系并不是那么紧密，建议从信号量的语义出发理解并使用信号量
        
        • 

• 锁在底层是怎么实现的？

  • 在实现锁的时候我们需要考虑的因素有哪些？
    
    • mutual exclusion – 基本要求
    • fairness
    • performance
  • 我们可以借助硬件开关中断的方式实现锁
    -
    - 但是开关中断的方法有寄个缺点：
    - this approach requires us to allow any calling thread to perform a privileged operation
    - the approach does not work on multiprocessors
    - turning off interrupts for extended periods of time can lead to interrupts becoming lost
  • 还有一系列的方式可以用来实现锁（自旋锁、队列锁）
  • linux系统使用的是一种叫做两段锁的锁

• 用锁实现thread-safe的数据结构
  

  • counter
    
    • 实现一个线程安全的计数器很简单
      
      • 
    • 效率更加高效的sloppy counter
      
      • 
      • 但是这种实现存在一个performance/accuracy tradeoff
      • 对上图代码一个小小的改进就是在get方法中首先对每个线程调用update方法，更近global counter之后再返回global counter，但是这种方法依然不能保证百分之百的精确（想一想为什么）
  • 链表
    
    • 
    • 我们在实现的时候需要特别小心流程控制处锁的获得和释放。
      

  A general design tip, which is useful in concurrent code as well as elsewhere, is to be wary of control flow changes that lead to function returns, exits, or other similar error conditions that halt the execution of a function. Because many functions will begin by acquiring a lock, allocating some memory, or doing other similar stateful operations, when errors arise, the code has to undo all of the state before returning, which is error-prone. Thus, it is best to structure code to minimize this pattern.

• 所以我们可以对以上的代码进行改进，得到一份更加鲁棒的concurrent链表的实现
  
  • 

持久化

I/O设备

• A Canonical Device
  
  • 
• The Canonical Protocol
  
  • 
  • polling
    
    • 
  • Lowering CPU Overhead With Interrupts
    
    • 
  • difference between interrupts and polling
    
    • if a device is fast, it may be best to poll; if it is slow, interrupts
    • If the speed of the device is not known, or sometimes fast and sometimes slow, it may be best to use a hybrid that polls for a little while and then, if the device is not yet finished, uses interrupts.
  • DMA
    
    • using programmed I/O (PIO) to transfer a large chunk of data to a device, the CPU is once again overburdened with a rather trivial task
    • without DMA（c表示从内存copy内容到设备）
      
      • 
      • with DMA
        
        • 
  • Methods Of Device Interaction
    
    • explicit I/O instructions（in and out）
    • memorymapped I/O
• 操作系统是怎么保证不同的设备插上之后就能即插即用的呢？
  
  • incorporate各种设备的驱动程序
  • 

Interlude: File and Directories

• virtualization of storage

file

• Creating Files
  
  • 
• Reading and Writing Files
  
  • read() and write()
• Reading And Writing, But Not Sequentially
  
  • 
  • 
• Writing Immediately with fsync()
  
  • 
  • 
  • you also need to fsync() the directory that contains the file foo. Adding this step ensures not only that the file itself
    is on disk, but that the file, if newly created, also is durably a part of the directory
• Renaming Files
  
  • atomicity of renaming
    
    • 
    • 
    • 在上面的例子中，如果我们能实现rename操作的原子性，我们就能实现文件更新操作的原子性
• Getting Information About Files
  
  • stat() or fstat()
  • 
• Removing Files
  
  • 

directories

• Making Directories
  
  • the format of the directory is considered file system metadata, you can only update a directory indirectly by, for example, creating files, directories, or other object types within it
  • 
• Reading Directories
  
  • opendir(), readdir(), and closedir()
  • 
  • 
• Deleting Directories
  
  • rmdir()
  • rmdir() has the requirement that the directory be empty (i.e., only has “.” and “..” entries) before it is deleted.

Hard Links and Soft Links

• Hark Link，两个文件指向同一个inode
  
  • 每个inode有一个reference count项，表示指向这个inode的硬链接的数目
  • can’t create one to a directory
  • can’t hard link to files in other disk partitions
• Soft Link，软链接文件是一种新类型的文件，里面存储的是指向的文件的完整路径
  
  • dangling reference problem
• 在*nix系统下创建hard link或者soft link，使用ln命令

怎么组织同一个目录下的文件或者子目录？
现代操作系统一般使用B树。

对于目录和纯文件来说，二者存储的信息有什么区别？
纯文件，存储的是文件的内容，对于目录，存储的是目录下的子目录和文件以及它们对应的inode number。

Very Simple File System（VSFS）

• 

  • S表示superblock
  • i表示inode bitmap（allocation structure，free list也是一种选择）
  • d表示data bitmap（allocation structure，free list也是一种选择）
  • D表示data block

• inode

  • 存储文件的metadata
    
    • 
  • 一个很重要的问题是怎么通过inode表示出此文件在磁盘上的存储位置？
    
    • The Multi-Level Index（ext2 ext3）
      
      • direct pointer（most files are small） + indirect pointer（support for big files）
    • extent（ext4）
      
      • An extent is simply a disk pointer plus a length (in blocks)
      • extent-based file systems often allow for more than one extent, thus giving more freedom to the file system during
        file allocation.（适合磁盘空间比较大的情况）
    • linked list
      
      • 

• Directory Organization

  • 
  • 

• Free Space Management

  • 
  • 

• Access Paths: Reading and Writing

  • /foo/bar, read it, and then close it

  • 

  • /foo/bar（之前不存在） wirte it, and then close it

  • 

• Caching and Buffering

  • static partitioning
    
    • Early file systems thus introduced a fixed-size cache to hold popular blocks. As in our discussion of virtual memory, strategies such as LRU and different variants would decide which blocks to keep in cache. This fixed-size cache would usually be allocated at boot time to be roughly 10% of total memory.
  • dynamic partitioning
    
    • many modern operating systems integrate virtual memory pages and file system pages into a unified page cache

Locality and The Fast File System

VSFS的效率太低了，我们需要更加快速的操作系统，Fast File System (FFS)。

• Organizing Structure: The Cylinder Group

  • The first step was to change the on-disk structures. FFS divides the disk into a bunch of groups known as cylinder groups (some modern file systems like Linux ext2 and ext3 just call them block groups)
  • 每个group cylinder的结构和VSFS中的一样
    
    • 
  • How To Allocate Files and Directories
    
    • placement of directories
      
      • find the cylinder group with a low number of allocated directories (to balance directories across groups) and a high number of free inodes (to subsequently be able to allocate a bunch of files
    • For files, FFS does two things
      
      • makes sure (in the general case) to allocate the data blocks of a file in the same group as its inode
      • it places all files that are in the same directory in the cylinder group of the directory they are in.
    • The Large-File Exception
      
      • The first twelve direct blocks were placed in the same group as the inode; each subsequent indirect block, and all
        the blocks it pointed to, was placed in a different group

• A Few Other Things About FFS

  • sub-blocks（to solve internal fragmentation）
    
    • 
    • 
  • parameterization
    
    • In particular, the problem arose during sequential reads. FFS would first issue a read to block 0; by the time the read was complete, and FFS issued a read to block 1, it was too late: block 1 had rotated under the head and now the read to block 1 would incur a full rotation
    • 
    • 对于现代的磁盘来说，这都不是问题
      
      • Fortunately, modern disks are much smarter: they internally read the entire track in and buffer it in an internal disk cache (often called a track buffer for this very reason). Then, on subsequent reads to the track, the disk will just return the desired data from its cache.
  • support for long file names（symbolic link就是从这开始出现的）

• summary for FFS

  •